Background. Metastasis to vital organs is the cause of death in a large majority of cancer patients. Although it is of eminent clinical importance to understand how systemic disease evolves, metastasis remains one of the least understood aspects of cancer progression. We still do not have the answers to many fundamental questions: Are metastasis founders a random selection of cells from primary tumors in which all cells have essentially equal metastatic ability? Or do specialized metastatic clones evolve, perhaps in intermediate sanctuary spaces like regional lymph nodes, and then proceed to colonize distant body parts? Are metastases formed late in tumor progression, by highly evolved and aggressive clones that are the winners of many years of selection within the primary tumor? Or can metastases already be formed early in tumor development, by less evolved tumor cells? If both scenarios exist, would such ?early? metastases behave differently from ?late? disseminating tumor cells? Method. We have developed a methodology to reconstruct a cancer's evolutionary history with great accuracy. Our method relies on the analysis of insertion/deletion mutations in hypermutable, non-coding polyguanine repeats. The lineage information encoded in these sequences is unusually rich: genotyping of only a few dozen repeats can outperform exome sequencing for phylogenetic reconstruction. Using polyguanine fingerprinting, we have recently shown that regional lymph node metastases are not the source of liver metastases in most colorectal cancers, in contrast to a widely held paradigm. Aims and Impact. Here, we propose to build upon these results and further elucidate critical events in the evolution of metastatic colorectal cancer. We will determine if lethal distant metastases in the lungs evolve from specialized metastatic clones that reside in the colonic lymph nodes. We have previously established that 65% of liver metastases are seeded directly from the primary tumor, but the anatomy of the gastrointestinal vasculature suggests that this percentage could be significantly lower for other distant metastases. If this hypothesis were confirmed, our understanding of the role of the lymphatics in colorectal cancer would be fundamentally altered. Second, we present preliminary data indicating that distant metastases whose evolutionary trajectory diverged from the primary tumor in early progression stages are considerably less aggressive than metastases that emerge in later stages. We propose to study and confirm this phenomenon in a large patient cohort. The successful outcome will be a simple, cost-effective test to predict long-term survival in a subset of patients with metastatic cancer. Since some patients can survive for years or even decades in spite of metastatic disease, while others die within weeks of diagnosis, such risk stratification would be of substantial benefit to both patients and their care providers.